1. Field of the Invention
The present invention relates to an unmanned vehicles for the transportation of parts and finished products in factory premises in general, and in particular to a self-contained unmanned vehicles capable of independent locomotion on a pre-determined trajectory or travel path on the basis of pre-determined geographical data.
2. Description of the Prior Art In connection with the significant progress made in recent years in the area of factory automation (FA), various types of unmanned vehicles have been developed for the automatic handling of parts and components etc. in factory, warehouse, and other locations. While, for example, magnetic induction type or optical induction type unmanned vehicles are in general use, the movement of these unmanned vehicles follows the signals of indicators located in the premises and designed to flag the vehicle trajectory in or on the ground.
The vehicles herein referred to as self-contained unmanned vehicles have been developed, and are capable of independently determining the nodal points to be traversed by searching for the optimum path on the basis of the target-point node indication and of automatically moving up to the target point node. The term "node" in this context implies a stop point, turn-off point, operating point or similar point and may also imply any point in which a change in the travel/movement condition of the unmanned vehicle in terms of its travel speed and forward direction of movement takes place.
Normally, a self-contained unmanned vehicle has a built-in memory into which are entered geographical input data about the travel path. This route mapping consists of coordinate data giving the coordinates of all and every nodal point and geographic data specifying the distance from the pre-determined travel path across the nodal points to the lateral boundaries (left and right side walls) of the trajectory. The geographical data, in particular, are set as the values for the distance from the pre-determined trajectory to the lateral boundaries of that trajectory at pre-defined intervals (several tens of millimeters) along said pre-determined trajectory, and entered as the memory input. As and when required, the appropriate data are read out from the memory and form the basic reference data for the automatic movement control of the self-contained unmanned vehicle until the target nodal point.
To facilitate the understanding of the present invention, a pulse motor of prior art will be described next, with reference to FIGS. 8 to 10.
The top part of FIG. 8 is a general front view of the unmanned vehicle 1A, with 2L being the left-hand drive wheel, 2R the right-hand drive wheel, 3L the motor for driving the left-hand drive wheel 2I, 3R the motor for driving the right-hand drive wheel 2R, 4L the pulse-encoder for detecting the rotational speed of the left-hand drive wheel 2L, 4R the pulse-encoder for detecting the rotational speed of the right-hand drive wheel 2R, and 5 & 5 are wheels. The wheels 5 & 5 are free to rotate with respect to their respective shaft cores and also free to rotate with respect to the shaft direction perpendicular to the respective shaft cores. 6L and 6R are ultrasonic signal transmitters/receivers (hereinafter referred to as ultrasound sensors) for detecting the distance to the left and right trajectory boundaries. 7A is a control device.
As shown in FIG. 9, said control device 7A has a central processing unit (CPU 8, a program memory 9A, a work memory 10A, an interface circuit 11 and a motor drive circuit 12. A program controlling CPU 8 is written into said program memory 9A, while the geographical trajectory data are written into the work memory 10A. These geographical data consist of coordinate data giving the coordinates of all and every nodal point and geographic data specifying the distance from the pre-determined travel path across the nodal points to the lateral boundaries (left and right side walls) of the trajectory.
FIG. 10 shows the geographical data MLA giving the distance to the left trajectory boundary side wall W written into the work memory 10A. Said geographical data MLA are composed of the data la1 - lan giving the distances to the left side wall from the pre-determined trajectory traversed by the unmanned vehicle 1A at every specified distance 1. These distance data la1 - lan consist of two bytes each. Similarly, the geographical data MRA which are not shown here and give the distances to the right side wall of the vehicle's trajectory are composed of individual distance data similar to the geographical data MLA described above. These data are written into the work memory 10A.
Item 13 in FIG. 9 is a communication device, receiving instructions (giving the target nodal points) received by wireless transmission from the central control station (not shown here). The instructions received by said communication device 13 are furnished to the control unit 7A.
This conventional self-contained vehicle 1A with above configuration has a CPU 8 for controlling the vehicular movement as described below in accordance with the program written into the program memory 9A.
If a target node is provided by the central control station, the CPU 8 will search for the appropriate trajectory on the basis of the geographical data written into the work memory 10A so as to determine the node(s) to be traversed on the way toward the target. In the interval between one node and the next, the CPU 8 reads out the appropriate distance to the left and right side wall boundaries at each pre-defined distance 1, on the basis of the geographical data MLA and MRA. The CPU 8 drives the motors 3L and 3R so that the vehicle moves at a constant speed along the pre-determined travel path across the consecutive sequence of nodal points. The CPU 8 measures the distances to the left and right side wall boundaries on the basis of the signals supplied from the ultrasound sensors at each pre-defined distance 1. It also measures the travel distance from the node traversed immediately beforehand on the basis of the signals supplied from the pulse encoders 4L and 4R. From these measurement results, the CPU 8 assesses whether or not the current travel position deviates from the correct travel position obtained on the basis of the geographical data, and corrects the position if a deviation has been detected. This ensures that the unmanned vehicle 1A constantly travels through the correct travel position to arrive at the target node (destination) by passing through the consecutive sequence of nodal points.
The conventional unmanned vehicle 1A described above gives rise, however, to the following problems (1)-(4). (1) When the measurement intervals for measuring the distance to the side walls W is shortened to increase the accuracy of travel, a vast memory capacity will be required since the necessary quantity of geographical data will be commensurate with the number of measurements performed.
Thus, for example, the number of geographical data required to traverse only the distance lo is lo/1 so that, if the distance to the side walls is given at a millimeter accuracy, a 2 byte (16 bit) memory capacity is required for 1 geographical data since the width of the trajectory is normally in the meter order of magnitude. If, by way of example, 1=10 mm and lo=5,000,000 mm, the memory capacity required for the geographical data will be EQU x x,500,000/10 =10.sup.6 (byte) (1).
If geographical data are available for both sides, the total memory capacity required for all geographical data will be double the memory capacity calculated in (1).
If, however, l assumes a value of 50 mm, the memory capacity will be 1/5th that of equation 1, but this will lead to uneven, jittery travel control with the impossibility of achieving smooth locomotion.
(2) Since the geographical data are expressed in terms of the side wall distance at each pre-defined interval 1 traversed by the unmanned vehicle 1A, it follows that there are no geographical reference data within this distance 1. It is therefore not possible to achieve a very fine travel position control.
(3). A large number of data are required to draw up the geographical data so that the preparation of these data is a laborious and time-consuming procedure. Since, furthermore, data are handled in the form of numerical data, it is clear that it is not easy for humans to make sense of these.
(4) As the environment surrounding &:he trajectory becomes complicated, and the number of nodal points increases, so the quantity of information to be entered as memory input increases. This not only makes the input operation a tedious and time-consuming task, but also leads to the problem that when attempting to display all node data, the display will be extremely difficult to view so that errors may be easily committed during data input if this display is used for reference. The input operation may therefore take considerable time.
Conventional self-contained unmanned vehicles thus require, as the above geographical data, the inputting of a large number of distance data giving the distance to the trajectory side walls at pre-defined path intervals, so that a large memory capacity is required and fine travel position control cannot be achieved. The compilation and entry of these geographical data is therefore tedious and time-consuming and the geographical data are difficult to make sense of.